removing adapter sequences

[morning latte]

Hello,

I am working with Illumina Hiseq data (100-bp PE). I am trying to remove adapter sequences using Trimmomatic. I've got adapter sequences from the sequencing core I used. But some of adapter sequences still remain after running Trimmomatic when I checked them using FastQC. Any suggestions would be great. Thanks.

[jimmybee]

Can you give us an example of what you ran and what you're getting as an output? We can't really help you unless we get some background....

[ahnguyen]

I am having essentially the same problem originally psoted above. I want to remove adapter sequences from Illumina 100 bp PE reads. I run the following with Trimmomatic:

java -classpath ~/Scripts/Trimmomatic-0.30/trimmomatic-0.30.jar org.usadellab.trimmomatic.TrimmomaticPE -phred33 -trimlog trim_2.log R1.fastq R2.fastq T1.fastq T1.unpaired.fastq T2.fastq T2.unpaired.fastq ILLUMINACLIP:adapters.fa:3:40:15 LEADING:3 TRAILING:3 SLIDINGWINDOW:4:15 MINLEN:15

My adapters.fa file includes the following (in addition to others):
>DNA_primers_1
AGGGAGGACGATGCGG
>DNA_primers_2
CCGCTGGAAGTGACTGACAC
>RNA_linkers_2
GTGTCAGTCACTTCCAGCGG

I ran FastQC prior to trimming the adapters and the over-represented sequences include
Sequence Count Percentage
CCGCTGGAAGTGAC... 115468 0.44
AGGGAGGACGATGC... 112267 0.427
CCGCTGGAAGTGAC... 109341 0.416
AGGGAGGACGATGC... 105312 0.401
CCGCTGGAAGTGAC... etc etc
CCGCTGGAAGTGAC...

After running Trimmomatic, run FastQC on the new fastq's and get the following for overrpresented sequences:
Sequence Count Percentage
CCGCTGGAAGTGAC... 109151 0.426
AGGGAGGACGATGC... 106128 0.414
CCGCTGGAAGTGAC... 102985 0.402
AGGGAGGACGATGC... 99644 0.389
CCGCTGGAAGTGAC... etc etc
CCGCTGGAAGTGAC...

Is this not surprising? I think a lot of the remaining adapter sequences are adapters linked to each other, so they are well represented in the first 10 bps.

- Andrew

[mastal]

Hi Andrew (ahnguyen),

I think the adapter sequences you are using (for example the 16 bp DNA_primers_1 and the 20 bp DNA_primers_2) are not long enough for Trimmomatic to recognise a match, given the thresholds you are using
(3:40:15, so 40 for palindrome clipping and 15 for simple clipping).

If you look at the Trimmomatic web page,

http://www.usadellab.org/cms/?page=trimmomatic

on the last paragraph of the section titled 'The Adapter Fasta', it explains that
'Each matching base adds just over 0.6' to the score, so even if your read matches the adapter sequence perfectly, it would score only 20 X 0.6 = 12.
You have set the threshold for simple clipping to 15, a score which none of your reads will reach, so trimmomatic will not recognize any of the reads as having the adapter sequence you want to trim.

[arcolombo698]

how do you create the adapters.fa file?
I have the same problem

[mastal]

The more recent versions of Trimmomatic include adapters.fa files for Illumina Truseq v2 and v3.

See the link to the Trimmomatic web page that I gave in the post above if you don't already have Trimmomatic installed on your computer.

Have a look at that, and then if you want to use other adapter sequences you can either add them to the file, or make your own file.